The Difficult

mesothelioma from the Wittenoom crocidolite mine produce a range of estimates dependent on the values of lag time and lung clearance rates. Although published data suggest a minimum lag period of around 20 years there is considerable spread of another 30 years to cover most cases. Also, the assumption of reduced toxicity due to clearance of fibres from the lung is speculative. de Klerk et al' have estimated future Wittenoom cases using a model based on lag time and exposure variables but their predictions are somewhat higher than Berry's. As the future occurrence of mesothelioma due to past exposure to asbestos is of great interest to government authorities, industry, workers, and the community, the accuracy of these estimates is of importance. An alternative predictive approach is proposed using the distribution of time lag period since first asbestos exposure derived from information obtained in occupational histories from the Australian mesothelioma statistics.2 Assuming a similar latency distribution in the Wittenoom cases collected in 1980-5, it is possible to calculate the total number of cases expected from the mine and the number expected in each time period. For example, to calculate the total number of mesotheliomas expected for an employment period 1940-4, use the two observed cases recorded from this period in 1980-5 and the midpoint of employment to yield a latency of 37 542-5 years. From the lag data in Ferguson et al2 prepare a graph of the cumulative percentage of cases v lag time since first exposure and use this to estimate that these two cases represent 13% (34% less 21%), hence predicting a total of 13 3. Calculation of the expected cases from the 1940-5 employees occurring in the time period 1991-5 (lag of 52-5-47-5 years) gives 11% (88% less 78%) of 13-3 cases or 1-3. Table 1 shows the overall data set. The uncertainties in such an approach is that it includes environmentally exposed cases and takes a simplified view that there has been a similar distribution of age, the specific occupations, and exposures during the period of operation of the mine and mill, all of which are known to be incorrect.34 Also it assumes compatibility between the Wittenoom group and the whole Australian population. The estimates (71) appear reasonable, however, when compared with the 59 cases reported to the Australian Register in 1986-90. These register records were not used in the calculation as they were not collected with the same rigour as the 1980-5 data and hence cases may have been missed and they are known to contain less detailed information on occupational history.' Table 2 gives a comparison of the estimates from the various methods. The proportional latency calculation method predicts a total of 525 mesotheliomas from the 6505 male and 411 female workforce and past residents; 366 cases will occur in the period 1986-2020, numbers will peak in the period 1991-1995, and tail off over the next 25 years. This fits comfortably in the range and distribution of most likely values presented by Berry. A ROGERS National Institute of Occupational Health and Safety, GPO Box 58, Sydney, NSW 2001, Australia